Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
4. LIQUIDITY CHARACTERISTICS, IMPLICIT INFORMATION OF ASSET PRICES AND MONETARY POLICY IN CHINA 1 Xiao WEIGUO2 Zhao YANG3 Yuan WEI4 Abstract This paper empirically analyzes the liquidity characteristics of asset price boom-busts and the implicit information of asset prices in China during the period 1998-2011. The results indicate that liquidity plays an important role in asset price cycles and asset prices contain specific information of future output and inflation. Housing price booms are more consistent with credit expansion than stock price. Likewise, housing price has stable positive connection with future output gap and inflation relative to stock price. Based on the above results, the Chinese monetary policy should intervene in asset prices misalignment when in need, even targeting at housing price as conditions permit. Besides, credit control should be adopted to restrain housing price effectively. Keywords: liquidity characteristics, asset prices, monetary policy,COBS approach JEL Classification: E21, E42, E52 1 Introduction After the American financial crisis, worldwide loose monetary policy stimulates the recovery of global economy, however, aggravating the global liquidity surplus. In China, the asset prices represented by stock price and housing price sustained a sharp rise in 2009. Huge liquidity played a crucial role in this process (Li and Deng, 2011). With the rapid development of capital market in China, the asset prices have gradually had tighter connection with real economy, financial stability and monetary policy. Asset prices fluctuation has already implied rich information of monetary policy decisions (Wu, 2007). The Chinese central bank (People’ s Bank of China, PBC) has 1 This study was presented at 2013 Global Business, Economics, and Finance Conference, Wuhan University, China, 9-11 May 2013. The research is supported by Major Program of National Social Science Fund of China (No. 12&ZD046), Humanities and Social Science Project of Ministry of Education of China (No. 11YJA790169), and China Postdoctoral Science Foundation Funded Project (No. 2012M521446). 2 Economic and Management School of Wuhan University. E-mail: [email protected]. 3 Wuhan Branch of the People’ s Bank of China. E-mail: Sunnyzhao@ whu.edu.cn 4 Corresponding Author: Yuan Wei, Economic and Management School of Wuhan University, Luojiashan, Wuhan, China, 430072. E-mail: [email protected] Liquidity Characteristics, Implicit Information of Asset Prices implemented tightening policy to curb housing price hikes since 2010. With liquidity tightening policies, the real economy grew sluggishly, but potential inflation pressure still existed. Therefore, the monetary authority of China fell into a dilemma. Since money and credit aggregates are still important contents in the conduct of Chinese monetary policy, it is essential to pay closer attention to liquidity characteristics in asset price cycles and to the information embedded in asset prices. Liquidity is mainly denoted by monetary liquidity (in the sense of monetary aggregates in various calibers and money structure) and banking system liquidity (in the sense of total assets or total liabilities of the banking system and their term structure) at the macro-level (MGCCER, 2008). After the American financial crisis, the traditional view ‘monetary liquidity results in asset bubbles’ has been developed further into ‘expansionary monetary policy causes asset bubbles’. The improved view emphasized the causative effect of monetary and credit expansion on asset prices (Luo, 2011). Many existing empirical results proved that monetary liquidity had significant effect on asset prices (Baks and Kramer, 1999; Congdon, 2005; Li and Deng, 2011; Liu and Yin, 2011), and credit expansion is the leading cause of asset price inflation (Bordo and Jeanne, 2002; Gerdesmeier et al., 2010; Guo, 2010; Wang, 2010). Compared with monetary liquidity, total credit has stronger impact on asset prices (Machado and Sousa, 2006). Several studies concluded that the relationship between liquidity and asset prices is not simple linear, i.e. the liquidity expansion is closely related to asset price booms, while the connection between liquidity contraction and asset price busts is not that so (Adalid and Detken, 2007; He et al., 2011). The relationship also depends on the environment of interest rate, inflation and output, etc. At present, the academia and global central banks’ practices have reached a consensus on the necessity that monetary policy should concern about asset prices. Nonetheless, whether, when and how should central banks intervene in asset prices are still puzzles. Two of the core issues are as follows: Whether or not asset prices imply macroeconomic information and, can asset prices forecast future output and inflation precisely(Wang et al.,2008; Luo,2011)? In light of the complexity of the asset prices’ movement, the existing researches have not provided satisfactory solutions. Empirical evidences from different countries during different periods indicated that there were various combinations among asset prices, inflation and output (Stock and Watson, 2003; Han et al.,2008). Due to different views on the information implied in the volatility of asset prices, current studies offer alternative monetary policies. If the predictive ability of asset prices on output and inflation is strong, asset prices transmission and general price transmission of monetary policy have an automatic balancing mechanism. Consequently, a central bank’s efforts to stabilize sufficiently long-term inflation could also additionally realize the asset prices/financial stability objective (Schwartz, 1995; Borio, 2005). Asset prices should be incorporated into monetary policy decisions, only to the degree that those movements affect inflation expectation (Bernanke and Geltler, 2001; Miskin, 2007). Monetary policy should apply flexible inflation target, leaning against asset prices to reduce the cost of cleaning up the mess. However, all previous financial crises indicated that asset bubbles were apt to occur in a low and stable inflation environment. Furthermore, to some degree, asset prices volatized independently Romanian Journal of Economic Forecasting – XVI (4) 2013 57 Institute for Economic Forecasting relative to output and inflation. The reality brought a big challenge to the traditional policy. There are many serious practical difficulties in monetary policy concerning asset prices. First, policymakers can hardly identify whether asset prices are driven by fundamentals or non-fundamentals (Kohn, 2006). Second, it is difficult to construct a stable broad price index including asset prices as the nominal anchor. Third, considering the gains and losses of counter-fighting potential bubbles, the monetary authority needs to judge the intervening timing and the intervening depth accurately (Trichet, 2009). Finally, financial innovation integrates the money market, the credit market and the capital market, breaking the automatic balancing mechanism between general price transmission and asset prices transmission (Assenmacher-Wesche and Gerlach, 2008). Unfortunately, existing researches failed to develop a universal monetary policy frame theoretically. In specific countries and during specific periods, the information implied in asset price movements was diversified in empirical analysis. Correspondingly, the choice of monetary policy dealing with asset prices should have strong pertinence. There are two main motivations for this paper. First, based on identifying asset price booms and busts, our analysis endeavors to shed light on the liquidity characteristics of asset price cycles in China, so as to provide useful precautionary alert signals for the monetary authorities concerning asset prices. Second, disclosing the information contained in the movements of asset prices contributes to clear-cut comprehensions of the relationship between asset prices and the real economy. More importantly, taking the macroeconomic relevant information into account facilitates the PBC’s policy conduct, i.e. should, or how the PBC responds to asset price movements? The remainder of this paper is organized as follows: Section 2 identifies asset price booms and busts with the COBS approach and then illustrates the liquidity characteristics of asset price boom-busts in China. Section 3 performs empirical analysis on the implicit information of asset prices after presenting the empirical model. Section 4 presents the main findings and the implications for the monetary policy in China. 2. The Liquidity Characteristics of Asset Price Boom-Busts 2.1 Identifying Asset Price Booms and Busts with the COBS Approach The Constrained Smoothing B-splines (COBS) approach was suitably introduced by Ng (2005) to identify the asset price booms and busts. The COBS approach considers asset returns as the function of fundamentals and underlines the extreme of asset prices misalignment, covering the deficiencies of defining asset price booms and busts as right and left 10% tails of the HP trend. In this paper, the COBS approach is employed to analyze empirically the stock price and housing price in China from 1998 to 2011, imposing the theoretical restriction that asset returns should be nondecreasing with the potential real GDP growth. The implementation of COBS in the statistical package R chose quadratic spline as roughness penalty and used AIC to 58 Romanian Journal of Economic Forecasting –XVI (4) 2013 Liquidity Characteristics, Implicit Information of Asset Prices search automatically for the controlling parameter. The substitute variables are as follows: stock returns (SR), housing returns (HR) and output growth are proxied by growth rate of Shanghai Sock Exchange A-share closing indices (the same period of last year = 100), national housing sales price indices (the same period of last year = 100) and growth rate of value-added of industry (the same period of last year = 100), respectively. All series are adjusted by CPI (the same period of last year = 100). Monthly data was collected from the CECI database. The results are shown in Figure 1. For prudential considerations, we took asset returns over 90% conditional quantiles as booms and asset returns below 10% conditional quantiles as busts. In Figure 1, stock and housing price booms are in dark grey, while busts are in light grey. There are five boom periods and two bust periods of stock prices, and four boom periods and three bust periods of housing price. The boom periods and bust periods are in accordance with public intuition, basically. Figure 1. The Booms and Busts of Stock Price (Left) and Housing Price (Right) .16 2.4 2.0 .12 1.6 .08 1.2 .04 0.8 0.4 .00 0.0 -.04 -0.4 -.08 -0.8 1998 2000 2002 2004 SP quantile=0.2 quantile=0.8 2006 2008 2010 quantile=0.1 quantile=0.5 quantile=0.9 1998 2000 2002 2004 quantile=0.1 quantile=0.5 quantile=0.9 2006 2008 2010 quantile=0.2 quantile=0.8 HP 2.2 The Liquidity Characteristics Starting with liquidity indicators, we compare the asset price boom-busts with monetary liquidity and banking system liquidity. Real M2 growth rate (RM2) and M1/M2 are proxies for monetary liquidity in the form of aggregate and structure, respectively. Similarly, real total domestic CNY loans (RL) and the ratio of long-term loans to total loans(LL/TL) are proxies for banking system liquidity. Relevant data orignated from PBC data streams. Figure 2 and Figure 3 describe the above four liquidity indicators in stock price and housing price boom-busts, respectively. We can observe the following three characteristics. i. Stock price booms or busts were not fully consistent with the period during which money and credit aggregates expanded fast or contracted fast (mostly lagged). For instance, in the case of stock price booms, the linkage was not clear, as in some cases real M2 growth rate rose and in other it declined. Romanian Journal of Economic Forecasting – XVI (4) 2013 59 Institute for Economic Forecasting ii. The housing price booms were remarkably consistent with the fast-expanding periods of liquidity aggregates. Unexpectedly, liquidity aggregates also expanded in housing price busts. iii. M1/M2 had powerful explanation on stock price boom-busts. M1/M2 and LL/TL were both better structural liquidity indicators of interpreting housing price boom-busts. Figure 2. The Liquidity Characteristics in Stock Price Boom-Busts .6 .5 .3 .4 .3 .2 .2 .1 .1 2.0 2.0 1.6 1.5 1.2 1.0 0.8 0.5 0.4 0.0 0.0 -0.5 -0.4 -1.0 1998 2000 2002 RM2 2004 2006 2008 SR 1998 2010 2000 2002 2004 SR M1/M2 2006 2008 LL/TL 2010 RL Figure 3. The Liquidity Characteristics in Housing Price Boom-Busts .6 .39 .38 .5 .37 .36 .4 .35 .4 .34.3 .3 .33 .2 .3 .2 .1 .1 .0 .0 -.1 1998 2000 2002 RM2 2004 2006 HR 2008 2010 M1/M2 1998 2000 2002 HR 2004 2006 LL/TL 2008 2010 RL 3. Empirical Analysis of the Implicit Information of Asset Prices Asset prices can influence the real economy via wealth effect, Tobin Q and balance sheet channels of enterprises and households, etc. A vast amount of literature on developed countries used asset prices as predictor of output gap and inflation. The empirical conclusions of some literatures demonstrated that some asset prices 60 Romanian Journal of Economic Forecasting –XVI (4) 2013 Liquidity Characteristics, Implicit Information of Asset Prices predicted either output gap or inflation effectively, which ensured the possible existence of specific economic relevant information in asset prices. 3.1 The Model Following Stock and Watson (2003), we build model (1) and model (2) to analyze empirically the predictive ability of asset prices on future output gap and inflation. ygapt + h = µ1 + α1 ( L ) ygapt + β1 ( L ) APt + γ 1 ( L) Z1t + ξ1,t + h (1) Inflation t + h = µ 2 + α 2 ( L ) Inflation t + β 2 ( L ) APt + γ 2 ( L ) Z 2t + ξ 2,t + h (2) where: α 1 (L ), β1 (L ), γ 1 (L ), α 2 (L ), β 2 (L ), γ 2 (L ) are lag polynomials, h represents steps ahead of independent variables, µ1 , µ 2 are constants, ygapt is output gap at time t, Infation t is inflation rate at time t, APt denotes stock price or housing price at time t, Z1t , Z 2t represent additional predictors of output gap and inflation at time t. The benchmark models (3) and (4) are also employed. By comparing the forecast performance of (1) relative to (3) and (2) relative to (4), we can examine the predictive ability of asset prices. The forecast performance is measured by the Theil inequality coefficient. ygap t + h = µ 1 + α 1 ( L ) ygap t + γ 1 ( L ) Z 1t + ξ 1,t + h (3) Inflation t + h = µ 2 + α 2 ( L ) Inflation t + γ 2 ( L ) Z 2t + ξ 2,t + h (4) Evidences from China indicate that, in addition to asset prices, consumer price, producer price and monetary growth, etc. are the main influential factors of output gap (Zhang, 2009). The main factors affecting inflation are stock price, housing price, excess liquidity, output gap, RMB exchange rate and interest rate (Huang et al., 2010;Zhang, 2008; Wang, 2008). Consequently, the candidate predictors in Z 1t include consumer price, producer price and monetary growth. The candidate predictors in Z 2t include excess liquidity, output gap, RMB exchange rate and interest rate. 3.2 Data We use monthly data of eleven series over 1998-2011 as sample. Table 1 summarizes the substitute variables and their sources. Nominal GDP and nominal interest rate are adjusted by CPI (December 1997 = 100) to obtain real values. Others are adjusted by CPI (preceding month = 100) to obtain real values. The steps ahead of independent variables are set to three months, six months and nine months. The fixed three-months lags is used for α 1 (L ) , β1 (L ) , γ 1 (L ) , α 2 (L ) , β 2 (L ) , γ 2 (L ) . For datadependent lag lengths, these lag polynomials contain at most three non-zero coefficients. In the process of auto-regression, we remove the insignificant coefficients gradually. Romanian Journal of Economic Forecasting – XVI (4) 2013 61 Institute for Economic Forecasting Table 1 The Substitute Variables and Source Data series Output gap Consumer price/inflation Producer price Excess liquidity RMB exchange rate Interest rate Stock price Housing price Monthly GDP Substitute variables The deviation of GDP series from the HP trend CPI (preceding month = 100) Source CEIC database CEIC database PPI (preceding month = 100) M2 growth rate - GDP growth rate (preceding month = 100) RMB effective exchange rate indices RMB one-year benchmark deposit interest rate Growth rate of Shanghai Stock Exchange A-share closing indices (preceding month = 100) National housing sales price indices (preceding month =100 ) The sum of monthly total retail sales of consumer goods, investment in fixed asset and net export (in logarithm form) Wind database CEIC database BIS database CEIC database CEIC database Sina Financial Statistics CEIC database 3.3 Emprical Results and Analysis The empirical results are listed in Table 2 and Table 3. Inspection of Table 2 and Table 3 reveals four facts. i. By introducing the stock price, predictive ability of output gap in model (1) outperformed model (3) in the first and the second quarter, while it was inferior to model (3) in the third quarter. After introducing the housing price, model (1) sufficiently improved its predictive ability of output gap relative to model (3) in the first and the third quarter, but performed worse in the second quarter. ii. Stock price was negatively correlated with the first quarter ahead forecast of output gap, but positively correlated with the second and the third quarter ahead forecast of output gap. Nonetheless, the housing price was positively correlated with forecast of output gap in each quarter. iii. By introducing the stock price, the predictive ability of inflation in model (2) outweighed model (4) in every quarter, especially in the third quarter. After introducing the housing price, model (2) sufficiently improved its predictive ability of inflation relative to model (4) in each quarter, especially in the first quarter. iv. Stock price was negatively correlated with the first quarter ahead forecast of inflation, but positively correlated with the second and the third quarter ahead forecast of inflation. Nonetheless, the housing price was positively correlated with forecast of inflation in each quarter. The results suggest that the forecasting of output gap and inflation based on stock price or housing price succeeds. However, there are some differences between stock price and housing price. The forecasts of output gap based on stock price worked well in short-term, while the forecasts of output gap based on housing price worked well in relative long-term. The predictive ability of stock price on inflation was strong in relative long-term, while the predictive ability of housing price on inflation was strong in short-term. This means that the Chinese stock price and housing price movements 62 Romanian Journal of Economic Forecasting –XVI (4) 2013 Liquidity Characteristics, Implicit Information of Asset Prices contain specific information on future output and inflation in one or the other period. Besides, the housing price had stable positive connection with future output gap and inflation relative to stock price. Table 2 Forecasting Performance of Asset Prices on Output Gap Akaike Performance improved information by model (1) relative to criterion benchmark model (3) Stock h=3 -0.0021(-2.1015**) 0.5831 0.3168 0.48% Price h=6 0.0025(5.3383***) 0.5345 0.3105 0.49% ** 0.5677 0.3211 -4.6% h=9 0.0029(2.2761 ) Housing h=3 0.0223(1.8949) 0.5514 0.2655 1.3% Price h=6 0.0189(2.4391**) 0.4902 0.2761 -2.5% h=9 0.0568(-2.8113***) 0.6118 0.2383 5.1% Note: t-statistics in ( ) and the absolute value of t-statistics should be higher than critical value 2. *** ,**,* represent 1%, 5% and 10% significant level, respectively. The Theil inequality coefficient of benchmark model (3) is 0.5316, 0.5468 and 0.5271 when the value of h is 3, 6 and 9, respectively. Steps ahead Estimated coefficients Adjusted R2 Table 3 Forecasting Performance of Asset Prices on Inflation Performance improved by model Steps 2 (2) relative to Estimated coefficients Adjusted R ahead benchmark model (4) Stock h=3 -0.0036(-2.0020**) 0.3559 1.2475 0.21% Price h=6 0.0046(2.3331**) 0.3624 1.3712 0.39% *** 0.5618 0.8964 7.1% h=9 0.0196(6.6917 ) Housing h=3 0.1297(3.6714***) 0.4748 0.4297 6.3% Price h=6 0.0446(1.8921) 0.3415 1.3543 0.62% h=9 0.0386(2.0153**) 0.3617 1.2306 0.49% Note: The Theil inequality coefficient of benchmark model(4) is 0.5633, 0.5881 and 0.4458 when the value of h is 3, 6 and 9, respectively. Akaike information criterion 4. Conclusions and Policy Implications The empiricial analysis of liquidity characteristics in asset price cycles is obviously highly stylized and concludes three features that may be of practical significance, including the disaccord between liquidity aggregates and stock price, the high consistency between expansion of liquidity aggregates and housing price booms, and the great explanation power of structural liquidity indicators on stock price and housing price. During special period (i.e. the Asian financial crisis in 1998 and the American financial crisis in 2008), money and credit aggregates in China were expanding against asset prices depression, which was different from the mature economies. In general, asset prices substantially downward worsened the balance sheet of commercial banks and then the ‘credit grudging’ behavior prevailed. Subsequently, Romanian Journal of Economic Forecasting – XVI (4) 2013 63 Institute for Economic Forecasting credit growth declined sharply in asset price busts. As for China, the credit desicion of commercial banks is not exactly market-oriented, but depends on the government need of macro-control (Yuan, 2012). Moreover, the monetery policy decision is affiliated to some degree to the fiscal policy with soft constraint (Qian, 2007). When government engages in counter-cyclical fiscal policies to stimulate economic recovery, the PBC injects huge liquidity into market through the banking system. Therefore, the asset price (especially housing price) surges following the expansion of money supply and credit. Additionally, compared with aggregate liquidity indicators, structural liquidity indicators provide more useful warning information for monetary policy concerning asset prices and potential risk of the financial system. On one side, money structure and term-structure of credit assets are apt to shorten in asset price booms, and vice versa. On the other side, liquidity structure contains richer information, including saving and investment dicisions of households, policy expectation and credit dicision of commercial banks, etc. Based on these results, it is necessary for the Chinese monetary policy to respond to asset prices promptly when structural liquidity indicators change fast. In the meantime, the rapid expansion of credit aggregate and short term sructure of credit assets become important indicators of possible housing bubbles in China. Implicit information embeded in asset prices shows that both stock price and housing price have specific predicitive power for the future output gap and inflation, while housing price has more stable positive connection with future output gap and inflation relative to stock price. This difference could be reflected by the history of Chinese stock market and housing market development. Stock and Watson (2003) argued that the predictive power of asset prices relied on the nature of shocks hitting on the economy, the degree of financial markets development and other institutional details differing across countries. In the past twenty years, housing investment and its pulling effect on upstream-downstream industries have made tremendous contributions to China’s double-digit economic growth. The housing industry was taken as one of the pillars of national economy. Therefore, the volatility of housing prices exerts a great influence on the macroeconomy. In contrast, the Chinese stock market had experienced several institutional reforms and it is still not mature. At most times, stock price could not reflect the true fundamentals, and then has weak predictive power on future output gap and inflation. Lv (2005) proved that stock price did not Granger cause real economic growth, although there was co-integrating relationship between stock price and China’s real economy. Owing to the implicit information embedded in asset prices, the Chinese monetary policy should intervene in asset price misalignments when in need, even targeting at housing price when conditions permit (e.g., completion of interest rate liberalization, more flexible RMB exchage rate system and more independent central bank). In addition, considering the closer relationship between housing price booms and credit expansion, the Chinese monetary policy should adopt strict credit control and supervision to restrain housing price effectively. 64 Romanian Journal of Economic Forecasting –XVI (4) 2013 Liquidity Characteristics, Implicit Information of Asset Prices References Adalid R. and Detken C., 2007. Liquidity Shocks and Asset Price Boom/Bust Cycles, European Central Bank, Working Paper, No. 732. Assenmacher-Wesche K. and Gerlach S., 2008. Financial Structure and the Impact of Monetary Policy on Asset Prices, Swiss National Bank Working Paper 2008/16. Baks K.. and Kramer C., 1999. Global Liquidity & Asset Prices: Measurement Implication and Spillover, IMF Working Paper 99/168, pp.1-33. Bernanke B.S. and Gertler M., 2001. Should Central Banks Respond to Movements in Asset Prices? American Economic Review, Vol.91, No.2, pp. 253-257. Bordo M.D. and Jeanne O., 2002. Monetary Policy and Asset Prices: Does ‘Benign Neglect’ Make Sense? International Finance, 5(2), pp.139-164. Borio C., 2005. Monetary and Financial Stability: So Close and Yet So Far? National Institute Economic Review, 192, pp.84-101. Congdon T., 2005. Money and Asset Prices in Boom and Bust, Institute of Economic Affairs, IEA Hobart Paper 153. Gerdesmeier D., Reimers HE. and Roffia B., 2010. Asset Price Misalignments and the Role of Money and Credit, International Finance, International Finance, 13(3), pp. 377-407. Guo W., 2010. Assets Price Fluctuation and Bank Credit: Theoretical and Empirical Analysis Based on the Perspective of Capital Constraint, Studies of International Finance, 4, pp. 22-31. Han X.H., Zheng Y.Y. and Wu C.M., 2008. Empirical Study of the Relationship between Inflation and Housing Prices, Journal of Tsinghua University (Science and Technology), 3, pp. 329-332. Huang Y.P., Wang X. and Hua X.P., 2010. What Determine China’s Inflation? Journal of Financial Research, 6, pp. 26-59. Kohn D.L.., 2006. Monetary Policy and Asset Prices, Speech at a European Central Bank Colloquium Held in Honor of O. Issing, Frankfurt, March. Li J. and Deng Y., 2011. On the Monetary Factor Behind the Housing Prices Increase: An Empirical Comparative Analysis of USA, Japan and China in the Bubble Period, Journal of Financial Research, 6, pp.18-32. Li Q., 2009. The Policy Connotation of Assets prices Fluctuation: Empirical Test and Index Construction, The Journal of World Economy, 10, pp. 25-33. Liu G. and Yin T., 2011. Research on the Shock Liquidity on Asset Market in China, Economy and Management, 9, pp. 58-63. Luo Z.Y., 2011. The Volatility of Asset Prices, Economic Cycles and Evolvement of Monetary Policy Regulation, Economic Perspectives, 3, pp.121-126. Lv J.L., 2005. Should China’s Monetary Policy Respond to the Change of Stock Price? Economic Research Journal, 3, pp. 80-90. Romanian Journal of Economic Forecasting – XVI (4) 2013 65 Institute for Economic Forecasting Machado, J.A.F. and Sousa J., 2006. Identifying Asset Price Booms and Busts with Quintile Regressions, Banco de Portugal Working Papers, No.8. Miskin F.S., 2007. Housing and Monetary Transmission Mechanism, Paper Presented at the Fed of Kansas City 31st Economy Policy Symposium, August 31-Semptemper 1. Ng P., 2005. A Fast and Efficient Implementation of Qualitatively Constrained Smoothing Splines, Proceedings of the 2005 International Conference on Algorithmic Mathematics and Computer Science. Qian X. A., 2007. Excess Liquidity and Monetary Policy, Journal of Financial Research, 8, pp. 15-30. Stock J.H. and Watson M.W., 2003, Forecasting Output and Inflation: The Role of Asset Prices, NBER Working Paper Series, No.8180. Schwartz A., 1995. Why Financial Stability Depends on Price Stability, Economic Affairs, Autumn. Trichet, J-C., 2009. Systemic Risk, Clare Distinguished Lecture in Economics and Public Policy, Clare College, Cambridge University. Wang H., Wang Y. and Fan C.L., 2008. On the Indicatory Function of Stock Price for the Monetary Policy, Journal of Financial Research, 6, pp. 94-108. Wang X. M., 2010. The Cyclical Relationship between Bank Credit and Assets price, Journal of Financial Research, 3, pp. 45-55. Wu G., 2007. Monetary Policy and Asset Prices: Classical Theory, Practice of the FED and Some Reflections, Nankai Economic Studies, 4, pp. 90-105. Yuan W., 2012. The Volatility of Asset Prices and the Choice of Chinese Monetary Policy, Doctoral Dissertation, Wuhan University. Zhang C.S., 2008. The Nature of Inflation Inertia in China and Its Implications on Monetary Policy, Economic Research Journal, 2, pp. 33-43. Zhang C.S., 2009. Estimating Output Gap: A Multivariate Dynamic Model Approach, Statistical Research, 7, pp. 27-33. 66 Romanian Journal of Economic Forecasting –XVI (4) 2013